17822535. ONLINE FAIRNESS MONITORING IN DYNAMIC ENVIRONMENT simplified abstract (International Business Machines Corporation)

From WikiPatents
Jump to navigation Jump to search

ONLINE FAIRNESS MONITORING IN DYNAMIC ENVIRONMENT

Organization Name

International Business Machines Corporation

Inventor(s)

Manish Kesarwani of Bengaluru (IN)

Pranay Kumar Lohia of Bhagalpur (IN)

Ramasuri Narayanam of Andhra Pradesh (IN)

Rakesh Rameshrao Pimplikar of Bangalore (IN)

Sameep Mehta of Bangalore (IN)

ONLINE FAIRNESS MONITORING IN DYNAMIC ENVIRONMENT - A simplified explanation of the abstract

This abstract first appeared for US patent application 17822535 titled 'ONLINE FAIRNESS MONITORING IN DYNAMIC ENVIRONMENT

Simplified Explanation

The abstract describes a method, computer program, and computer system for online fairness monitoring using a trained machine learning model with protected attributes.

  • Dataset with protected attributes and model data is received.
  • Entry with maximum reward selected based on reward probability.
  • Bias detection in the model towards protected attributes based on changes in reward probabilities exceeding a threshold.

Potential Applications

  • Fairness monitoring in online platforms.
  • Ensuring non-discriminatory outcomes in machine learning models.

Problems Solved

  • Detecting bias in machine learning models.
  • Promoting fairness and equality in decision-making processes.

Benefits

  • Increased transparency in algorithmic decision-making.
  • Mitigation of potential discriminatory practices.
  • Enhanced trust and accountability in machine learning systems.


Original Abstract Submitted

A method, computer program, and computer system are provided for online fairness monitoring. A dataset having one or more entries with one or more protected attributes and data corresponding to a trained machine learning model is received. An entry having a maximum reward is selected based on a reward probability associated with the entry. A determination is made as to whether bias has developed in the trained machine learning model toward one or more of the one or more protected attributes based on a change to the reward probability or a distribution of reward probabilities exceeding a threshold value.

Retrieved from "http://wikipatents.org/index.php?title=17822535._ONLINE_FAIRNESS_MONITORING_IN_DYNAMIC_ENVIRONMENT_simplified_abstract_(International_Business_Machines_Corporation)&oldid=37091"